Is anything real? Testing "fake" LM308s

[derevaun]

A few weeks ago I stumbled upon an Ebay listing for some LM308 "solder pulls" for pretty cheap: https://www.ebay.com/itm/386473025200  When I watch-listed it, the seller sent an even cheaper offer, with this note:

How about 25% off, eh?
What do you call a fake noodle? An impasta!
Thanks for checking out our new store!

Obviously, clearly, almost-came-out-and-said-it, fakes, right? Who could resist. I ordered a set, breadboarded a Rat, and waited.



When they arrived, I tested a few types of chips:

• the maybe fakes from this Ebay listing
• a Motorola LM308H from that "No China Bullshit" git on Ebay
• a few AMD LM308H from a small stash I lucked into last year
• a couple LM308N from Retroamplis
• an LM108, PIFfed from Laundryroom David


• a definite fake LM308N from Ebay several years ago (it has a date code that matches a lot of dubious Ebay listings)

...and...

• LM301
• CA3130
• LM307 (internally compensated)
• LM741 (internally compensated)

...and...

• A few УД1408А (the supposed Soviet 308 equivalent)

All of the 308s (and the 108), except the fake 308, measured pin 5 not-connected and M-Ohms across 1 and 8.

I auditioned them all on the breadboarded Rat, by strumming a low chord and hammering some high notes, with the 30pF capacitor in and then out. This cap-in-or-out difference is what I listened for.

With the cap in, they all sounded pretty consistent. But removing the cap:

-the LM301 introduced some gritty, screechy artifacts on chord and note decay
-the CA3130 introduced some messy decay and a gritty, busy self noise
-the LM307 and LM741 had no difference, cap in or out, but the 741 was a lot more sensitive to touching the capacitor
-the  definite fake LM308 behaved like the LM307: no diference (it already sounds fine in a DOD 250 that doesn't have a 1-8 capacitor, also pin 5 is connected)

All of the LM308s, from the Motorola to the ones in the Ebay listing, and the УД1408А, responded to cap removal the same way: they got brighter and grittier, and a little more ragged with note decay. I didn't notice a salient difference between any of them. But the 741, 301, 3130, and 307 were each clearly different, in their own way, with regard to the external cap-in-or-out.

I also tried some a couple NTE 938 (308 replacements) that I had squirreled away, and they behaved like the 308s.

I'm not fully convinced that the Ebay "solder pull" LM308s are genuine, but they behave the same with the cap-in-or-out test, and I don't have a reliable way to compare them otherwise (I don't trust my ear to be neutral after a power-down-chip-swap-power-up interval). The real test is going to be just playing them over time and possibly developing affinity for some over others. But declare wholesale that these are fakes? I can't.

In the name of science art, I ordered 3 more sets. Some of them have identical date codes, which is another hint that they may be fakes. But all of them behaved the same as the other 308s with the cap-in-or-out test.

To be clear, I do not vouch for the Ebay seller or any suggestion of authenticity of any LM308s in the listing. I just can't distinguish the difference between the ones I got and some LM308s that I'm pretty sure are legitimate. I suspect there are other, more dispositive tests for a difference? I'd love to hear about other ways to test them. But, the more time and effort I invest in this, the less it seems to matter as a musician.

[jimilee]

....But, the more time and effort I invest in this, the less it seems to matter as a musician.

Mojo parts are great, but it really comes down to this. Any musician worth their salt has lost some hearing and couldn't tell you the difference. I'm kidding, take care of your ears, hearing aids are expensive.


Sent from my iPhone using Tapatalk

[Aleph Null]

At the risk of getting philosophical, what's the matter with them being fake if they sound right, other than maybe having paid too much for them?

[jessenator]

At the risk of getting philosophical, what's the matter with them being fake if they sound right, other than maybe having paid too much for them?

Preach on, my dude. I'm one of those who is definitely in that camp of ears over badge/brand/mojo. I also at least like to try the "original" or as close as we can get. Give it all a fair shake if you can, but in the end go with what sounds good to you.


Derevaun: SUPERB white up, sir! Hopefully sometime today I'll be able to break open the PiF you sent and I can go board up a bog standard Rat. I'll do that before popping them into my Fat Rat.

[jessenator]

Haven't done any significant test, other than "yup, that works"

Turns out my UD1408A is the anomaly! Here's the orientation that worked for each (derevaun's UD on the far left)


printing wouldn't all show up from the same angle of light

[blearyeyes]

Reality is a simulation. Still, wouldn’t it be great if you could get verified original functional components? I built a little Chinese signal generator kit. It cost me $5 delivered. The chip it uses XR2206 costs more than that. The fake chip in the kit makes signals. But higher frequencies are VERY distorted and unusable. The real chip will run @ over 12v where the fake one looses it. So is the fake junk as good as the real stuff? NOPE. It’s fake, recycled, or a reject from the factory. There is a video somewhere showing how to tell fake from real. The tonnage of electronics delivered to China to be recycled is ginormous. They are sorta good at sanding off and reprinting chips. Personally I’m going through a faze where I want to make gooder stuff with gooder parts but hey, I would like the best simulation possible.

[jwin615]

I'll add this
The Moto's in foam came from Jameco last week. The ones in the green holders came from eBay last summer(no fake Chinese bull.....)
They share the same production codes and look identical.
Is anything real?
Also, the Jameco PDIP LM308s fail resistance tests. They are fake. I only bought 3 but emailed them Friday night.

[jwin615]

Another less blurry photo
The 3 from eBay have 3 different date codes
The 7 from Jameco share the same 3 codes across them

The Nat Semi logo on the pdips is all wrong.  It should be fatter in the middle of the N, not uniform thickness

[blearyeyes]

Like this?

[jessenator]

Like this?

That's the one, yup

https://www.dibsplace.com/design/ICLogos.htm#N

[blearyeyes]

Great resource! I missed the link first time around...

[jessenator]

Great resource! I missed the link first time around...

np!
Took me a minute to re-find it, but there's this one as well. Not as high res photos here, but some others not in the first link:
https://rtellason.com/semiconductorlogos.html

[derevaun]

Going on Chuck's post on the PPCB forum, I tested the Ebay LM308s for current. They measured between 260 - 360 microAmps, in the same range as the legit ones. So maybe they are legit. They sound fine to me, so it doesn't matter a whole lot.

The LM308N I got from Jameco is definitely fake. Actually it's mixed up with another fake LM308N that I got on Ebay a few years ago, but they both don't need the 1-8 capacitor. One is currently in a stock DOD250, where it sounds killer The other I included in the test above.

I've ordered NE555s from Jameco for a couple decades, for a class. They've worked fine for our uses (blinking LEDs, Atari Punk Console, PWM) but it's been a rogue's gallery of weird/sloppy markings. Probably factory rejects or similar, and they probably would fail some test for high precision, but I haven't had a single one not work for low-stakes uses.

I guess the pinch point is that we expect a particular kind of magic from the LM308. It's certainly beyond my ear's ability to nail down, but there's always gonna be a mystique that close enough isn't quite close enough 

FWIW, it looks like the Ebay seller reduced the price to $8.00 for 4, plus shipping. I absolutely do not vouch for the seller or anything you might receive from them, but my experience was good and perhaps others of us like to live dangerously. Or, for science! Here's the listing:

https://www.ebay.com/itm/386701906664?hash=item5a093ad2e8:g:C9oAAOSw0-NlsP61

[jwin615]

Anyone notice I accidentally took a pic of the 3080s I bought and not the 308s?
Lol

[blearyeyes]

It’s a miracle any build works. I didn’t notice…3080, 308 all the same to me.